WELD DATA

Citation Author(s):
Yang
Liu
Submitted by:
Enpei Guo
Last updated:
Fri, 08/30/2024 - 10:41
DOI:
10.21227/m58z-y515
License:
0
0 ratings - Please login to submit your rating.

Abstract 

      

The weld classification dataset contains a variety of high-resolution images that capture a wide range of weld types and conditions common in industrial environments. The dataset includes samples from a variety of welding processes, such as arc welding, laser welding, and resistance spot welding, representing a variety of materials, including steel, aluminum, and titanium alloys. Each image is carefully labeled by experienced welding inspectors, classifying welds according to quality (e.g., acceptable, defective), type (e.g., butt, fillet, lap) and specific defects (e.g., porosity, cracks, non-fusion). The dataset features welds in different orientations and under different lighting conditions to simulate real-world inspection scenarios. In addition, it combines challenging situations such as welds with subtle defects, complex geometry, and subject to surface contamination or oxidation. This comprehensive dataset not only serves as a powerful benchmark for evaluating weld classification algorithms, but also facilitates the development of models that can handle the complexity and variability in industrial welding applications.

Instructions: 

import numpy as np

import pandas as pd

import matplotlib.pyplot as plt

from pyts.image import MarkovTransitionField

for i in range(1):

 # data=pd.read_csv("WD%i-H.csv"%(i+1),header=None)

 data = pd.read_csv("124.csv", header=None)

 

 X =data.values.tolist()

 # MTF transformation

 mtf = MarkovTransitionField(image_size=10)

 X_mtf = mtf.fit_transform(X)

 # Show the image for the first time series

 plt.figure(figsize=(10, 10))

 plt.imshow(X_mtf[0], cmap='rainbow', origin='lower')

 #plt.title('Markov Transition Field', fontsize=18)

 #plt.colorbar(fraction=0.0457, pad=0.04)

 plt.tight_layout()

 plt.axis('off')

 plt.subplots_adjust(top=1, bottom=0, right=1, left=0, hspace=0, wspace=0)

 plt.margins(0, 0)

 # plt.savefig ("WD1.png"%(i+1))

 plt.savefig("WD1.png")

 plt.show()

 i=i+1

Comments

Where is the dataset?

Submitted by Daniel Griffin on Fri, 11/22/2024 - 12:49

Dataset Files

    Files have not been uploaded for this dataset